Abstract
Recent studies point to power consumption becoming the major design constraint in exascale computing systems. Current scientific benchmarks, such as LINPACK, only evaluate high-performance computing (HPC) systems when running at full throttle, i.e., 100 % workload, resulting in more of a focus on performance than on power and energy consumption. In contrast, efforts like SPECpower evaluate the energy efficiency of a server at varying workloads. This is analogous to evaluating the fuel efficiency of an automobile at varying speeds. However, the applicability of SPECpower to HPC is limited at best.
Given the absence of a scientific benchmark to evaluate the energy efficiency of HPC system at different workloads, we propose GBench (short for Green Benchmark), a methodology to evaluate the energy efficiency of supercomputers and enable a more rigorous study of energy efficiency in HPC. We use LINPACK as a case study and demonstrate the efficacy of our methodology by identifying application parameters impacting performance and providing a systematic methodology to vary the workload of LINPACK.
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00450-012-0218-0/MediaObjects/450_2012_218_Fig1_HTML.gif)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00450-012-0218-0/MediaObjects/450_2012_218_Fig2_HTML.gif)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00450-012-0218-0/MediaObjects/450_2012_218_Fig3_HTML.gif)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00450-012-0218-0/MediaObjects/450_2012_218_Fig4_HTML.gif)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00450-012-0218-0/MediaObjects/450_2012_218_Fig5_HTML.gif)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00450-012-0218-0/MediaObjects/450_2012_218_Fig6_HTML.gif)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00450-012-0218-0/MediaObjects/450_2012_218_Fig7_HTML.gif)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00450-012-0218-0/MediaObjects/450_2012_218_Fig8_HTML.gif)
![](http://media.springernature.com/m312/springer-static/image/art%3A10.1007%2Fs00450-012-0218-0/MediaObjects/450_2012_218_Fig9_HTML.gif)
Similar content being viewed by others
Notes
Exascale systems are predicted to consume about 67 megawatts (MW) of power [2].
References
Barroso LA, Hölzle U (2007) The case for energy-proportional computing. Computer 40(12):33–37
Bergman K, Borkar S, Campbell D, Carlson W, Dally W, Denneau M, Franzon P, Harrod W, Hill K, Hiller J, Karp S, Keckler S, Klein D, Lucas R, Richards M, Scarpelli A, Scott S, Snavely A, Sterling T, Williams RS, Yelick K, Kogge P (2008) Exascale computing study: technology challenges in achieving exascale systems
Feng X, Ge R, Cameron KW (2005) Power and energy profiling of scientific applications on distributed systems. In: IEEE IPDPS. doi:10.1109/IPDPS.2005.346
Fast Correlation-Based Filter (FCBF) Solution Software (2003) Available at http://www.public.asu.edu/~huanliu/FCBF/FCBFsoftware.html
Ge R, Feng X, Song S, Chang H, Li D, Cameron KW (2010) PowerPack: energy profiling and analysis of High-Performance systems and applications. IEEE Trans Parallel Distrib Syst 99(2). doi:10.1109/TPDS.2009.76
High performance LINPACK (HPL) (2008) Available at http://www.netlib.org/benchmark/hpl
HPC Challenge Benchmarks (2003) Available at http://icl.cs.utk.edu/hpcc
Hsu C, Feng W, Archuleta JS (2005) Towards efficient supercomputing: a quest for the right metric. In: IEEE IPDPS HPPAC workshop
Kamil S, Shalf J, Strohmaier E (2008) Power efficiency in high performance computing. In: 2008 IEEE international symposium on parallel and distributed processing, Miami, FL, USA, pp 1–8
NAS parallel benchmarks (1992) Available at http://www.nas.nasa.gov/Resources/Software/npb.html
Performance Application Programming Interface (PAPI) (2011) Available at http://icl.cs.utk.edu/papi
Press WH, Flannery BP, Teukolsky SA, Vetterling WT (1992) Numerical recipes in C: the art of scientific computing, 2nd edn. Cambridge University Press, Cambridge
Song S, Ge R, Feng X, Cameron KW (2009) Energy profiling and analysis of the HPC challenge benchmarks. Int J High Perform Comput Appl 23(3):265–276
SPECpower benchmark (2008) Available at http://www.spec.org/power_ssj2008
Tan TZ, Goh RSM, March V, See S (2009) Data mining analysis to validate performance tuning practices for HPL. In: 2009 IEEE international conference on cluster computing and workshops
The Top500 list (1993) Available at http://top500.org
Yu L, Liu H (2003) Feature selection for high-dimensional data: a fast Correlation-Based filter solution. In: The twentieth international conference on machine learning
Author information
Authors and Affiliations
Corresponding author
Additional information
This project was supported in part by the US National Science Foundation (NSF) via grant CCF-0848670.
Rights and permissions
About this article
Cite this article
Subramaniam, B., Feng, Wc. GBench: benchmarking methodology for evaluating the energy efficiency of supercomputers. Comput Sci Res Dev 28, 221–230 (2013). https://doi.org/10.1007/s00450-012-0218-0
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00450-012-0218-0